[email protected] wrote: > Yes that will work. I am not sure if the link layer fault detect will > work correctly so you might need to run Spanning Tree also. Something > that can be a issue is if say you have 4 links and one is running 24mbit > modulation and the rest are 54, your going to have issues with the slow > link. If possible I would use a radio board that can take all your > radios and bond them, presenting you with a single ethernet with the > bonded capacity.
For what it's worth, I talked to a buddy today who does quite a bit of switching stuff (especially with iSCSI related stuff), and he recommended against running spanning tree. He had some questions on the types of radios that I'd be sending these LACP 802.3ad packets through (to make sure that they passed them through to the switch on the other side), but he said that if all of that was kosher, then he'd just stick with LACP features and avoid adding STP to complicate things. He also recommended using the *dynamic* LACP features, rather than static features, as the static features were really designed for legacy devices and did primitive load balancing like round robin (which could cause problems in the lower modulation scenario that you gave). Thanks for your feedback. If people are interested, I'll post the solution that I find works best for me. -------------------------------------------------------------------------------- WISPA Wants You! Join today! http://signup.wispa.org/ -------------------------------------------------------------------------------- WISPA Wireless List: [email protected] Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
